Chapter 9: Deep Learning Fundamentals¶

Build neural networks from scratch, then master PyTorch—CNNs for images, RNNs for sequences, and regularization for production.

Metadata¶

Learning Objectives¶

• Implement a neural network from scratch with NumPy (forward pass, backpropagation)
• Understand activation functions, loss functions, and gradient descent
• Build and train models with PyTorch (nn.Module, DataLoader, optimizers)
• Apply regularization: dropout, weight decay, early stopping, batch normalization
• Build CNNs for image classification
• Build RNNs and LSTMs for sequence prediction
• Diagnose underfitting/overfitting and tune hyperparameters

What's Included¶

Notebooks¶

Scripts¶

• deep_learning_toolkit.py — NeuralNetScratch class, activation functions, plotting utilities

Exercises¶

• 5 exercises with solutions (in solutions/ branch)

SVG Diagrams¶

• 3 visual diagrams for network architecture, backpropagation, and CNN structure

Read Online¶

You can read the full chapter content right here on the website:

• 09.1 Introduction -- Neural networks from scratch, backpropagation, gradient descent
• 09.2 Intermediate -- PyTorch tensors, autograd, training loops, regularization
• 09.3 Advanced -- CNNs, RNNs, transfer learning, image classification capstone

Or try the code in the Playground.

How to Use This Chapter¶

1. Clone and install dependencies

git clone https://github.com/luigipascal/berta-chapters.git
cd berta-chapters
pip install -r requirements.txt

2. Navigate to the chapter

cd chapters/chapter-09-deep-learning-fundamentals

3. Launch Jupyter

jupyter notebook notebooks/01_introduction.ipynb

Created by Luigi Pascal Rondanini | Generated by Berta AI